🩺 MedSimple: AI-Powered Clinical Text Simplification

MedSimple is a mobile application designed to bridge the health literacy gap. It uses optical character recognition (OCR) and large language models (LLMs) to scan jargon-dense medical discharge summaries and instantly translate them into patient-friendly, 5th-grade reading level bullet points.

⚠️ The Problem

Patients often leave the hospital confused by complex discharge summaries, lab reports, and clinical notes. Medical jargon like "Myocardial Infarction" or "Idiopathic Hypertrophy" creates a barrier to understanding, which can lead to poor health outcomes, medication non-adherence, and increased readmission rates.

💡 The Solution

MedSimple empowers patients by putting a medical translator in their pocket. By simply snapping a photo of their medical paperwork, the app processes the text and provides a clear, concise, and actionable summary. It highlights critical terms and formulates specific questions the patient should ask their doctor during follow-up visits.

🛠️ Tech Stack & Architecture

This project is built using a modern, scalable, and cross-platform mobile architecture:

• Frontend: React Native & Expo (Cross-platform iOS/Android development)
• Hardware Integration: expo-image-picker and native camera permissions
• Text Extraction: OCR.space Public API for high-speed document parsing
• Generative AI: Google Gemini 2.5 Flash API (Cloud-based LLM inference)

🚀 Features

• Live Camera Scanning: Native device camera integration for on-the-fly document capture.
• Intelligent Simplification: Leverages the reasoning capabilities of Gemini 2.5 to simplify text while maintaining clinical accuracy.
• Actionable Insights: Automatically flags severe conditions or medications and separates them into a "Follow-up Questions" section.
• Clean UI: Built with an accessible, high-contrast, "Apple Health" inspired interface.

⚙️ Getting Started (Local Development)

Prerequisites

• Node.js (v18+)
• npm or yarn
• Expo Go App installed on your physical mobile device
• A Google AI Studio API Key

Installation

1. Clone the repository:

   git clone [https://github.com/YourUsername/MedSimple.git](https://github.com/YourUsername/MedSimple.git)
   cd MedSimple

2. Install dependencies:

   npm install

3. Environment Setup: Create a .env file in the root directory and add your Gemini API key:

   EXPO_PUBLIC_GEMINI_API_KEY=your_api_key_here

4. Start the Expo Server:

   npm start

5. Run the App: Scan the QR code printed in the terminal using the Expo Go app on your phone.

🧠 Future Roadmap

As a researcher focused on robust and equitable multimodal clinical intelligence systems, the next phases of this project include:

1. On-Device Inference: Migrating from the cloud-based Gemini API to an on-device local model (e.g., Gemma 2B via MediaPipe) to guarantee absolute data privacy and HIPAA compliance, ensuring no patient data ever leaves the phone.
2. Multilingual Support: Expanding the LLM prompt structure to translate complex English medical texts into clear, simple Urdu and other languages to support non-native speakers.
3. Acoustic Integration: Adding a speech-to-text pipeline to allow users to record doctor's instructions and receive the same simplified breakdown.

---

Disclaimer: MedSimple is a proof-of-concept tool designed for educational purposes. It is not a substitute for professional medical advice, diagnosis, or treatment.